next up previous contents index
Next: Setting up a LexGram Up: What is a LexGram Previous: The fundamental phrase structure

The head-driven chart parser

The LexGram parser implements the fundamental rule of LexGram in the following manner. Now, substrings are represented by their string positions PhonFrom and PhonTo.

Data structures:

'Prove Goal Sequent':
map a goal sequent g on a goal table t

'Fill Arguments':

extend the goal table t of goal sequent g = PhonFrom,PhonTo ; Slash |- GoalCategory

with projections of the constituent h = PhonFrom1,PhonTo1 ; HeadSlash, HeadCategory:

The LexGram generation algorithm works essentially in the same way. However, as resources it takes a list of essential elements of the input semantics (the 'guide') instead of the input string. For further discussion of head-driven chart parsing see the paper by van Noord [11].


next up previous contents index
Next: Setting up a LexGram Up: What is a LexGram Previous: The fundamental phrase structure
Esther Koenig-Baumer
6/23/1998